Over the last few weeks we’ve blogged a couple of times on Decision Optimization
- The Customer Journey to Decision Optimization
- Ongoing Decision Improvement Drives Decision Excellence
- Don’t Just Sit There, Experiment!
Decision-making is often a contentious subject. As humans, we think the way we make decisions is part of our “secret sauce”, our value to the organization. As a result, there are generally lots of opinions about how decisions should be made. Many rely on gut feelings and sniff tests. There’s often lot of talk of things “everyone knows”. Historical approaches – current practices – are often referred to as “best practices” even though no-one has ever really challenged them. And, of course, any decision made manually by multiple people is made inconsistently.
All this makes finding the best decision-making approach an argumentative activity. Proposals are written, comments are collected, memos are sent and meetings are held. Agreement often seems to require a lowest common denominator approach or an acceptance of inconsistency.
Today there is an alternative. Instead of arguing about decision-making, you can model decision-making. You can develop what are called Decision Requirements Models using the Decision Model and Notation standard. These models are based on a simple idea – that decisions you take repeatedly can be described as questions with allowed answers and that these decisions can be broken down to show the sub-decisions they depend on – and the sub-sub decisions etc. How these smaller decisions are made (knowledge) and the information they need (input data) can also be captured. A simple notation allows business owners, data scientists, operations teams and IT all to share an understanding of how a decision is made.
You can also use influence diagrams to show you how your decisions influence your business outcomes, using them to tie your data to the decisions you make with that data, to the key metrics you track and how these metric drive profit, loss and overall success. Influence Diagrams can evolve into Decision Impact Models that define mathematical relationships. Decision Impact Models can be executed to predict the impact of different actions on your business to identify the decisions that will meet your objectives.
With these models, there’s no longer any argument about how you decide or how you tell good decisions from poor ones. Its all clearly defined in the model. The role of these diagrams in your journey to optimal decisions is discussed in our new paper.
So, stop arguing and start modeling. Your decisions will thank you.